In general, hydrocarbon contaminations on EM specimen surfaces are always unavoidable even though TEM analysts care more about specimen contaminations and SEM analysts care less. For instance, after a few hours in air or a night in a microscope with ‘poor’-vacuum, a new layer of organic molecules can be formed on biological TEM specimen surfaces. Old designs of TEMs had given rise to the characteristic appearance of contamination deposits; however, contamination has been significantly reduced by improving the column design and vacuum systems in modern microscopes.

During TEM observation, the contaminated area is normally a crown, annular spot, or needle with a diameter nearly equal to that of the illumination size. The thickness of the surface contamination varies with time. However, the actual shape of contamination structure and contamination rate depend mainly on a couple of factors:
        i) The vacuum level of the microscope.
        ii) The size of incident electron bean (see in Figure 4392a).
        iii) The electron probe current density.
        iv) The sticking coefficient for hydrocarbons on the specimen.
        v) The existing contamination within the microscope vacuum system.
        vi) The history of the specimen such as specimen preparation, handling of the specimen or specimen holder, back-streaming of oil from a diffusion-pumped ion milling system, chemicals used during electrolytic thinning, cleaning, and exposure to hydrocarbon vapors. For instance, after hours in air, a new layer of hydrocarbons can form on the specimen. Note that, in modern microscopes, most of the contamination is caused by surface diffusion of adsorbed hydrocarbon molecules to the irradiated area.
        vii) Charging effect. The charging effect on the contamination becomes dominant for small beam sizes due to generation of secondary electrons.

Figure 4392a. Schematic illustration of the shape of formed contamination structures: (a) Contamination needle with small beam sizes (< 20 nm), e.g. in CBED and STEM modes, when the electron probe is stationary, (b) Medium beam sizes, (c) Contamination film with large beam sizes, e.g. with parallel beam in TEM, or scanned area in STEM or SEM.

For SEM, at low beam current densities, the thickness (t_c) of the contamination film is given by,
         ------------------------------- [4392a]
                 ------------------------------- [4392b]
                 ------------------------------- [4392c]
where,
         S -- Stopping power (∝E^-0.8).
         E -- The dissipated energy.
         q -- The accumulated charge density (=jt if the beam is stationary, =I_pT/a if the beam scans).
         j -- The current density.
         t -- The irradiation time.
         T -- The frame time.
         a -- The scanned area.

Equation 4392a indicates that the contamination rate is higher under low-voltage SEM condition than under "high"-voltage SEM systems. Table 4392 lists some experimentally measured specimen contamination rates (dz_c/dt) depending on accelerating voltage (E₀) current density (j) and vacuum pressure (P).

Table 4392. Experimentally measured specimen contamination rates (dz_c/dt).

At high beam current densities in TEM and SEM observations, the growth rate of contamination film (for scanning beam) or needle (for stationary beam) saturates because all the molecules that reach the irradiated area by diffusion are polymerized immediately. Therefore, the growth rate of hydrocarbon only depends on the irradiation time t rather than the beam current density j. On the other hand, the stronger contamination appears on the side region of the scanned area because most molecules cannot reach the central area before depositing.

Contamination rate can practically be evaluated by EELS technique in TEM, backscattered electron signal in SEM and elastic scattering method in STEM. In modern TEMs, the contamination rates is generally well below 1 nm³min^-1. The contamination sources in TEM are mainly:
        a) Condensation of hydrocarbons and polymerization of organic molecules (e.g. from the atmosphere and vacuum pumps) at the irradiated area.
        b) Mobile adsorbed molecules on the specimen.
        c) Components decomposed or vaporized from the specimen (causing specimen damage) due to irradiation by the electron beam.

The partial pressure of hydrocarbon (HC), back-streaming of oil from the rotary pump and of silicon oils from the diffusion pump, the grease of vacuum seals, and fingerprints on TEM sample holder and specimen induce HC contamination in the EMs (electron microscopes) specimen chamber. This HC diffuses along the specimen surface into the beam-irradiating area, where they become polymerized. The polymerization dose has been measured as 1.6 mC/cm² and the surface diffusion coefficient as 55,000 nm²/s at 18 °C [1]. Carbon-metal bonds can be strong while C–C bonds are even stronger. Non uniform HC contaminants can distort the final image of the specimen. Note that the contamination induced by thermal diffusion is normally larger than direct deposition of organic molecules from the vacuum. All these contaminants are pinned by electron radiation when arriving at the irradiated area.

These contaminants on the TEM specimen surfaces limit a number of applications, e.g. HRTEM, low energy EDS analysis and EELS analysis. Various methods can be used to reduce or remove contaminations:
        i) Using ACD or cool-traps. In this way, the partial pressure of hydrocarbons and other gases (e.g. oil vapor) can be reduced in the specimen chamber. Note that the residual gas in the vacuum can be checked and analyzed by a mass spectrometer.
        ii) Coating non-carbon films (e.g. thin metal layers) on both surfaces of the carbon-contained TEM specimen.
        iii) Cleaning:
        iii.a) Plasma cleaning, UV exposure, electron beam flooding (or so-called beam shower) in TEM or argon (Ar) ion milling can be employed to remove the hydrocarbons and residual contaminants from the TEM specimen surfaces.
        iii.b) The specimen is heated up (e.g. ~30 °C) above the room temperature before or during TEM observation.
        iii.c) Destroy the adsorbed hydrocarbon molecules by ion-beam irradiation or by sputter coating of conductive layers.
        iii.d) Washing the specimen in pure methylene, methanol or ethanol (e.g. for a night).
        iii.e) Leaving the specimen in the microscope overnight and then the contaminants can fully or partially be desorbed in the vacuum.
        iii.f) Bake the microscope with the specimen in it.
        iii.g) The amount of contaminants in the specimen chamber can be decreased by gas-discharging.
        iii.h) Spray the specimen with N₂ or Ar through a capillary during the electron irradiation.
        iv) Specimen cooling.
        v) For EELS and EDS measurements, a diffusion barrier can be formed by beam-induced contamination to protect the interesting area from contamination. As shown in Figure 4392b, the hydrocarbon contamination barrier are formed before official analysis. That is, the elemental mapping in the interesting area  is performed after the diffusion barrier is formed. Therefore, the cracked diffusible contaminants cannot reach the interesting area.

        vi) Optimized vacuum. However, note that a "correct" vacuum alone does not solve the contamination problem if the specimen itself is neither clean nor prepared inside an ultra-high vacuum specimen chamber.
        vi.a) The TEM operates under ultrahigh vacuum (e.g. ~ 3 x 10^-8 Torr), obtained by a dry vacuum-pumping setup with both turbo pumps and ion pumps for oil-free vacuum. In general, the better the vacuum, the less the contamination. As shown in Figure 4392c (a), in most modern TEMs, the electron gun, top lenses, and specimen chamber are maintained at ultra-high vacuum by an ion pump, while the viewing screens and photographic chamber are maintained at a lower vacuum, which is referred to as high vacuum, by either a diffusion pump or a turbomolecular pump. This vacuum level is backed by a mechanical (rotary) pump. However, some TEMs have lower vacuum in the specimen chamber as shown in Figure 4392c (b). In fact, the main difference between the two types of microscopes is whether or not the specimen contamination is reduced by ion pumps.

        vi.b) Check and fix air leaks in the vacuum.
        vi.c) Use of a diffusion pump oil of low vapor pressure.
        vi.d) Viton O-rings and apiezon grease are employed instead of silicon grease for vacuum seals.
        vi.e) Polytetrafluoroethylene (PTFE) or polyimides are used if plastic material is employed for insulation.
        vi.f) Flood the microscope column and the specimen chamber with dry nitrogen.

Again, some of the methods above cannot directly address the sources of specimen-born contamination, while cleaning process can directly reduce or eliminate the contamination source. However, the cleaning procedure is different from case to case because the sticking coefficients for hydrocarbons depend on the materials in the TEM specimen, resulting in different contamination rates and different amount of hydrocarbon on the specimen.

In contrast, uniform HC contaminants can prevent or reduce beam damage, sputtering, and etching of the specimen.

[1] J. S. Wall. Scanning Electron Microscopy (1980) I in: O. Johari (Ed.), SEM Inc. Chicago, p.99.